Hi all, On Monday, 25 November 2013 19:30:51 CEST, Jan Kundrát wrote:
some people are working on making the Trojitá's website redesign. The initial version which I have been shown use various data from 3rd-party servers, like some CDN for JS delivery, GitHub's social buttons showing how many people have forked the repo, or Google's font service. I should probably note that even the current version includes SourceForge's code for donations via PayPal. In this mail I would like to ask for your opinion about whether you mind the possibility of tracking by these 3rd parties.
I do mind about the possibility of tracking.
I have mixed feelings about this -- on one hand, privacy-conscious users can set up an appropriate block list themselves if they mind the additional tracking.
Well, I already do this indeed. But quite often, websites have limited functionality if I disable JavaScript, and some websites even manage it to be completely unusable without JavaScript enabled, even though the website most important functionality is showing some text. That is stuff I really would like to prevent happening to the Trojita website. About fonts: I don't really care what font the text is in, as long as I can read it. So for me, it's not needed to fetch some nice font from Google's servers.
Linking to a CDN for static contents saves the bandwidth in the long run
Although that might be true, I personally don't really care about that. Being tracked as little as possible is more important to me than saving a few kilobytes on bandwidth. An even easier way to save bandwidth is not having a fancy website at all and instead just have a relatively simple website, like Trojita has now.
and integrating with 3rd party services like Ohloh and GitHub provides some value for the users (or for the project) -- a nice and not-so-obtrusive widget showing the latest commits to make it clear that the project is well alive, etc.
I don't know how other people do it, but if I encounter a website of a project I never saw before, I always go check out the repository to see what's happening. So what I do is click on a (for example) GitHub link myself and checkout what's happening. A widget showing the latest commits wouldn't be really necessary for me.
On the other hand, most of this can be made without depending on any 3rd party at all, either via some server-side scripting, or perhaps by proxying some of the contents via our own infrastructure (if the upstream provider actually allows this usage). Of course, everything *can* be done, but there's also the issue of manpower; I have better things to do than writing a post-update git hook which puts a couple of recent entries into a .json file to be displayed by JS on clients, etc.
The issue of manpower is a good point. However, if we keep the site as simple as possible (like it is now), a lot of manpower is not needed. If the functionality of the website stays roughly the same, I don't see why we would have a lot of extra work on maintaining the website.
So, which way shall we go -- shall we go with the masses and embrace the web blink and don't hesitate to link to third parties, or shall we insist on do-everything-ourselves and prevent any possibility of the 3rd-party tracking?
So summarized, my opinion is like this: if we keep things simple, there is no need to do a lot of work on the website and/or to depend on other websites. Best regards, Caspar Schutijser
